Chemical recycling of plastic waste to oil followed by subsequent steam cracking into light olefins will become the main route for reduction of the escalating plastic waste crisis in the coming decade. Plastic waste contains a significant amount of contaminants in the form of heteroatomic compounds containing O, N, S, Cl and Br. Steam crackers have strict limits on these contaminant concentrations to ensure a high quality product and protect the internal equipment of the steam cracker. As such the pyrolysis oils need to be treated before being send into the cracker. A new emerging technology is that of supercritical water pyrolysis, this is the thermal decomposition of plastic waste under the presence of supercritical water. This technology shows promise to produce a lighter type of pyrolysis oil and remove a fraction of the contaminants during the process. In this project general research will be done in the decomposition behaviour of plastic waste and its contaminants in supercritical water. To this end the reaction network generator Genesys will be expanded to include reactions with polymers. A machine learning model will be developed to estimate polymer properties. The constructed model shall be validated with experiments and finally the calculation time of the model will be reduced aided by machine learning models. This should give significant insight into plastic waste decomposition under supercritical water opening paths for its industrial application and optimization.